Table 11.

Sources of cash holding increases

CFICFFDebt issuanceDividends
Dependent variable(1)(2)(3)(4)
LMR–0.005 (−0.53)–0.034 (−1.34)–0.021 (−1.10)0.008** (2.17)
LMR#FFF0.002 (0.22)0.059** (2.17)0.010** (2.27)–0.016** (−2.77)
FFF–0.034 (−0.59)–0.376* (−1.81)–0.021 (−0.88)0.057** (2.72)
Size(t1)–0.000 (−0.00)–0.127*** (−3.31)–0.016** (−2.80)0.035*** (7.92)
Growth(t1)0.020*** (3.44)0.065*** (4.18)0.014 (1.48)0.004** (2.79)
Tangibility(t1)–0.325 (−1.41)–0.577* (−1.89)0.054** (2.15)–0.011 (−0.46)
NWC(t1)0.187** (2.33)0.033 (0.21)–0.151*** (−3.19)0.155*** (4.21)
Cash(t1)0.962*** (3.66)0.235 (0.62)–0.176** (−2.70)0.230*** (5.19)
Other ultimate owner0.004 (0.44)0.042 (1.61)0.028 (1.40)–0.000 (−0.03)
Inflation(t1)–0.750** (−2.46)–2.044*** (−2.96)0.530 (1.31)–0.568** (−2.90)
GDP growth(t1)–0.163 (−1.34)–0.257 (−0.63)0.092 (0.47)0.671*** (5.36)
ln(GDP p.c.)(t1)0.322** (2.52)1.539** (2.61)0.069 (0.47)–0.388*** (−3.47)
Firm fixed effectsYesYesYesYes
Industry-Year fixed effectsYesYesYesYes
Country specific year trendYesYesYesYes
Clustered standard errorsCountry-levelCountry-levelCountry-levelCountry-level
Observations27,91827,85624,32026,628
Radj20.0840.0510.0270.063
Note(s):

This table presents the results of OLS regressions of mechanisms behind the increases in cash holdings in family firms, compared to non-family firms. The analyses are conducted using interaction terms between labor market regulations and family firms (LMR#FFF). The dependent variables are Cash Flow from Investment Activities divided by Total Sales (t1) (Column 1), Cash Flow from Financing Activities divided by Total Sales (t1) (Column 2), Long Term Debt Issuance divided by Total Assets (t1) (Column 3), and Dividend Payout divided by EBIT (t1) (Column 4). All variables are defined in Table A1 in the  Appendix. All independent variables except LMR, FFF, and Other Ultimate Owner are lagged by one period. All regression specifications include firm fixed effects, industry-year fixed effects, and country specific year trends. The t-Statistics in parentheses are based on robust standard errors, clustered at the country level. *, **, *** denote statistical significance at the 10, 5, and 1% levels, respectively

or Create an Account

Close subscription notice
Close access options